209: - MR. SPRUCE'S BOTANICAL EXCURSION 
path—they were leaf-stalks of an Aroideous plant, possibly a Dracontium, 
called, in lingoa géral, Jararáca-tajá, i. e. snake-plant, and exactly re- 
sembled in their markings the venomous Jararáca. By searching in 
the wood we found a few young plants, but no flowers. Here, too, 
I was interested to fall in with the 4jaré, whose leaves are much used 
for killing fish: it is a pretty half-shrubby plant, with purple papi- 
lionaceous flowers, and leaves almost as silvery as those of Alchemilla 
alpina, I have heard it called, also, Cunambi, but this seems a general 
name for plants of which the same use is made. According to Martius, 
a Euphorbiaceous plant, the Phyllanthus Conami of Swartz, possessing 
similar properties, is called Cunabi in lingoa géral. 
It was here that our men were employed in constructing the toldas, 
there being close at hand great plenty of the Pindoba palm (the stem- 
less variety of Attalea compta, Mart.), the fronds of which suited their 
purpose. In the vicinity of Santarem and Obidos, the fronds of the 
Pindoba are universally used for thatching, as those of the Buss and 
Ubim are around Para. When growing, the pinne spread out hori- 
zontally on each side of the midrib; but, for thatching, those of one 
side are bent over between the interstices of those on the other, and 
thus a sort of long tile is made, the midrib of the frond forming its 
upper edge. A roof constructed of several layers of these is impene- 
trable by any rain. In preparing our toldas, the pinnæ were not 
bent to one side, but the fronds being laid together, so that each 
was half covered by the adjacent one, the pinnæ, crossing at right 
. angles, were then interwoven into a sort of mat. Three fronds 
. in breadth, and two or three in thickness, formed a perfectly water- 
|... proof tolda. 
. Shortly after starting this morning, we passed on the right an 
igarapé coming from the Lago Itapecurü. It is well to remark here 
that these igarapós all bear the same name as the lake from which they 
flow, and, as will hereafter be seen, the Trombétas is truly a river of 
lakes. Beyond the igarapé Itapecurü, the river gradually veers round 
— to W.N.W.; it then turns further, and assumes a N.W. direction, 
until the rounding of a point brings us to the mouth of the Rio Iripixy. 
Shortly, however, before arriving here, we pass the Furo Caxoiry, on 
the left, which leaves the Amazon nearly opposite the upper entrance 
. to the Lagoa Grande. I learnt that the Rio Iripixy had not a very 
long course, and that it also took its rise ina lake. Directly opposite 
* 
